Petitioner, David Ronald Chandler, was sentenced to death for murder in furtherance of a continuing criminal enterprise under 21 U.S.C. § 848(e)(1)(A). We affirmed Petitioner’s murder conviction and death sentence on direct appeal. United States v. Chandler, 996 F.2d 1073 (11th Cir.1993). Petitioner then filed a section 2255 petition challenging his conviction and sentence on several grounds. Among other things, Petitioner, invoking the Sixth Amendment, claimed that he received ineffective assistance of counsel — during the sentencing phase of trial — because his trial counsel failed to investigate and to present character witnesses. The district court rejected Petitioner’s claims and denied relief. We affirm.1
*1310BACKGROUND
Petitioner ran an extensive marijuana growing and distribution operation in Northern Alabama. In January 1990, Petitioner had offered Charles Ray Jarrell, Sr., one of Petitioner’s marijuana couriers, $500 to eliminate Marlin Shuler, a suspected informant; Jarrell has said he thought Petitioner was joking. On 8 May 1990,2 Petitioner saw Shuler at Jarrell’s house. Petitioner warned Jarrell that Shuler was going to cause them trouble and stated: “You need to go on and take care of him and I still got that $500.” Jarrell understood Petitioner to be referring to Petitioner’s earlier offer to pay Jarrell should he eliminate Shuler. Petitioner left; Jar-rell and Shuler spent the morning drinking heavily. The two men then drove to a lake for some target practice with two guns. During target practice, Jarrell turned a gun on Shuler, shot him twice, and killed him. Jarrell went to and informed Petitioner that he had killed Shuler; the two men returned to the scene and disposed of the body. Jarrell asked for, but did not receive, the $500.
In 1991, a nine-count indictment charged Petitioner with various drug, continuing criminal enterprise, and conspiracy offenses, including procuring Shuler’s murder in furtherance of a continuing criminal enterprise.
Petitioner retained Drew Redden, a prominent Alabama criminal defense lawyer, to defend him at trial.3 Redden actively pursued acquittal, especially on the charge of procuring a murder.4 In preparation, trial counsel, among other things, observed the trial of a codefendant and reviewed material from the defense attorneys of the other defendants to the CCE charge. He consulted a jury selection expert. And trial counsel interviewed at least 67 witnesses in the Piedmont and Esom Hill area, Petitioner’s small community. Believing his client not to be a true drug kingpin, counsel also spent time trying to find the “real” drug kingpin. Trial counsel, throughout the case, was in frequent contact with Petitioner, Petitioner’s brother, and Petitioner’s wife.
To contest the murder charge, trial counsel introduced evidence at trial to show the weaknesses in the Government’s case: that is, trial counsel attacked the idea that Petitioner had in reality caused, on the pertinent day, Shuler to be killed. A history of animosity existed between Jarrell and Shuler as a result of Shuler’s former marriage to Jarrell’s sister. Shu-ler had abused his ex-wife and mother-in-law (Jarrell’s sister and mother respectively), which provided Jarrell with his own motives for killing Shuler. Jarrell — on an earlier occasion and for his own reasons — ■ had actually attempted to kill Shuler: Jar-rell had placed a gun to Shuler’s head and pulled the trigger, but the loaded gun had just not gone off. In addition, trial counsel stressed that the key Government witnesses in this case, including Jarrell himself, testified in exchange for lesser sentences. Furthermore, Jarrell, over time, had made inconsistent statements about Petitioner’s responsibility for the murder: stating that he (Jarrell) did not do it; stating that he (Jarrell) alone did it inten-*1311tionaUy because of personal animosity; admitting that he (Jarrell) did it but claiming it was an accident; and, at last, implicating Petitioner.
Nevertheless, the jury convicted Petitioner on all nine counts of the indictment, including the murder charge. The jury implicitly found that the Government proved beyond reasonable doubt that Petitioner had offered to pay, and induced, Jarrell to kill Shuler. A separate death penalty sentencing hearing on the murder count was held the next day.
At sentencing, the Government alleged three statutory aggravating factors: (1) that Petitioner had intentionally engaged in conduct resulting in the death of another, (2) that Petitioner procured the killing of another for money, and (3) that Petitioner committed the murder after substantial planning and premeditation. The Government offered no new evidence at the sentencing phase and relied on the evidence it had presented at the guilt phase.
Defense counsel did present evidence as well as arguments for mitigation, among other things, stressing residual doubt.5 He reminded the jury — using a stipulation about the date of the death of the victim— that a taped statement, made by Petitioner about having to kill somebody (a tape the jury had requested to review at the guilt-stage deliberations), was made three months after the murder of Shuler and did not indicate that Petitioner was talking about Shuler. Trial counsel also entered into evidence two other stipulations: (1) that Petitioner had no prior criminal record, and (2) that Jarrell, the actual killer, and Jarrell’s son (who was also implicated in the murder) would never be prosecuted for Shuler’s murder. Both of these latter stipulations were mitigating factors as a matter of law under the pertinent statute. In addition, trial counsel called as character witnesses and presented to the jury the humanizing testimony of Petitioner’s wife and mother.6
Given that the evidence was unconverted that Jarrell (not Petitioner) had actually killed the victim, trial counsel argued again at sentencing that the evidence was not absolutely conclusive about whether Jar-rell, especially in his drunken state, was truly induced by Petitioner when Jarrell shot Shuler. Trial counsel pointed to evidence that Jarrell earlier, in November 1989, had — completely independent of Petitioner — put “a pistol to the nose of [Shu-ler] and pull[ed] the trigger intending to *1312kill him”: the gun had misfired. Trial counsel reiterated the independent malice Jarrell harbored for the victim as a result of their personal history. Trial counsel asked the jury how Petitioner could have motivated Jarrell at the time of the killing, when Jarrell (after Petitioner had left) had consumed “twenty-three beers on that date, twenty-three beers before he shot the man.”7 Trial counsel stressed that imposing the death penalty in this case would be “cruel and unusual punishment” and a “tremendous mistake ... considering every circumstance of this case.”
The jury, however, found that the first two aggravating factors existed and unanimously recommended that Petitioner be sentenced to death. The district court, Judge Hancock, did so.
After exhausting his direct appeals, Petitioner moved to vacate his convictions and sentence under 28 U.S.C. § 2255 and moved for a new trial in accordance with Fed.R.Crim.P. 33 on many grounds, including ineffective assistance of counsel at sentencing. In the light of Petitioner’s claims, the district court, Judge Hancock, conducted a series of evidentiary hearings.
On the ineffectiveness claim, Petitioner asserted that counsel was ineffective for failing to investigate and to present character witnesses at the sentencing hearing. At the section 2255 evidentiary hearing, Petitioner presented 27 witnesses who testified to specific good acts by Petitioner.8 Petitioner also presented the testimony of defense counsel Redden.
Judge Hancock found that the mitigating value of the proffered witnesses was undercut on cross-examination: (1) the good character evidence related to a time remote from that of Petitioner’s crimes; (2) many of the witnesses were ignorant of Petitioner’s criminal activities altogether, showing an ignorance of Petitioner’s character; and (3) all of the witnesses showed a strong bias in favor of Petitioner.9
The district court then considered the aggravating circumstances and found that the jury had determined that two aggravating factors existed based on a “particularly egregious crime”: encouraging another, dependant upon him, to kill a police informant and then assisting in disposing of the body. The district court — “[wjeighing this weak character evidence against the strong aggravating evidence that the jury accepted” — concluded that prejudice was not proved. Therefore, the district court denied relief on the ineffectiveness claim.
DISCUSSION
Petitioner argues that he received ineffective assistance of counsel during the sentencing phase of his trial. We review Petitioner’s claim of ineffective assistance de novo. Williams v. Head, 185 F.3d 1223, 1227 (11th Cir.1999). To succeed on a claim of ineffective assistance, Petitioner must show both incompetence and prejudice: (1) “[PJetitioner must show that ‘counsel’s representation fell below an objective standard of reasonableness,’ ” and (2) “[Pjetitioner must show that ‘there *1313is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.”’ Darden v. Wainwright, 477 U.S. 168, 106 S.Ct. 2464, 2473, 91 L.Ed.2d 144 (1986); accord Williams v. Taylor, — U.S. -, 120 S.Ct. 1495, 1511, 146 L.Ed.2d 389 (2000). Petitioner says that his trial counsel was incompetent because his trial counsel failed to investigate and to present character evidence. And, Petitioner says that — but for his trial counsel’s failure to investigate and to present the evidence — a reasonable probability exists that the jury would not have voted for a death sentence.
SOME PRINCIPLES GOVERNING PERFORMANCE
To aid courts in assessing claims of ineffective assistance under the Sixth Amendment, the Supreme Court and this court, particularly sitting en bane, have set out certain principles and presumptions.10 We have recognized that, given these principles and presumptions, “the cases in which habeas petitioners can properly prevail ... are few and far between.” Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir.1995) (en banc). A review of the principles and presumptions seems appropriate.11
I. The standard for counsel’s performance is “reasonableness under prevailing professional norms.” Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 2065, 80 L.Ed.2d 674 (1984); accord Williams v. Taylor, — U.S. -, 120 S.Ct. 1495, 1511, 146 L.Ed.2d 389 (2000) (most recent decision reaffirming that merits of ineffective assistance claim are squarely governed by Strickland). The purpose of ineffectiveness review is not to grade counsel’s performance. See Strickland, 104 S.Ct. at 2065; see also White v. Singletary, 972 F.2d 1218, 1221 (11th Cir.1992) (“We are not interested in grading lawyers’ performances; we are interested in whether the adversarial process at trial, in fact, worked adequately.”). We recognize that “[r]epresentation is an art, and an act or omission that is unprofessional in one case may be sound or even brilliant in another.” Strickland, 104 S.Ct. at 2067. Different lawyers have different gifts; this fact, as well as differing circumstances from case to case, means the range of what might be a reasonable approach at trial must be broad. To state the obvious: the trial lawyers, in every ease, could have done something more or something different. So, omissions are inevitable. But, the issue is not what is possible or “what is prudent or appropriate, but only what is constitutionally compelled.” 12 Burger v. Kemp, 483 U.S. 776, 107 S.Ct. 3114, 3126, 97 L.Ed.2d 638 (1987).
II. The burden of persuasion is on a petitioner to prove, by a preponderance of competent evidence, that counsel’s performance was unreasonable. See Strickland, 104 S.Ct. at 2064; see also Williams, 120 S.Ct. at 1511 (“[DJefendant must show that counsel’s representation fell below an objective standard of reason*1314ableness.”) (internal citations and quotations omitted). The petitioner must establish that particular and identified acts or omissions of counsel “were outside the wide range of professionally competent assistance.” Burger, 107 S.Ct. at 3126; see also Strickland, 104 S.Ct. at 2064-65 (stating that petitioner must show “counsel’s representation fell below an objective standard of reasonableness” — that is, that counsel’s performance was unreasonable “under prevailing professional norms ... considering all of the circumstances”).
III. “Judicial scrutiny of counsel’s performance must be highly deferential.” 13 Strickland, 104 S.Ct. at 2065. We must avoid second-guessing counsel’s performance: “[I]t does not follow that any counsel who takes an approach we would not have chosen is guilty of rendering ineffective assistance.” 14 Waters, 46 F.3d at 1522 (en banc). Nor does the fact that a particular defense ultimately proved to be unsuccessful demonstrate ineffectiveness.
IV. Courts must “indulge [the] strong presumption” that counsel’s performance was reasonable and that counsel “made all significant decisions in the exercise of reasonable professional judgment.” Strickland, 104 S.Ct. at 2065-66; accord Williams v. Head, 185 F.3d 1223, 1227-28 (11th Cir.1999) (presuming counsel rendered effective assistance). Thus, counsel cannot be adjudged incompetent for performing in a particular way in a case, as long as the approach taken “might be considered sound trial strategy.” Darden v. Wainwright, 477 U.S. 168, 106 S.Ct. 2464, 2474, 91 L.Ed.2d 144 (1986). Given the strong presumption in favor of competence, the petitioner’s burden of persuasion — though the presumption is not insurmountable — is a heavy one.15 Kimmelman v. Morrison, 477 U.S. 365, 106 *1315S.Ct. 2574, 2586, 91 L.Ed.2d 305 (1986); see also Williams, 120 S.Ct. at 1511 (“[T]he defendant must show that counsel’s performance was deficient.” (quoting Strickland, 104 S.Ct. at 2064)).
V. The reasonableness of a counsel’s performance is an objective inquiry.16 See Darden, 106 S.Ct. at 2474 (noting that counsel’s performance did not fall below “an objective standard of reasonableness”); see also Williams, 120 S.Ct. at 1511 (same); Darden, 106 S.Ct. at 2474 (noting that “there are several reasons why counsel reasonably could have chosen to rely on” the defense that he did (emphasis added)); United States v. Fortson, 194 F.3d 730, 736 (6th Cir.1999) (determining — without district court findings or even evidentiary hearing — that defendant had not overcome presumption of effective assistance because court “[could] conceive of numerous reasonable strategic motives” for counsel’s actions at trial). And because counsel’s conduct is presumed reasonable, for a petitioner to show that the conduct was unreasonable, a petitioner must establish that no competent counsel would have taken the action that his counsel did take.17 See Waters, 46 *1316F.3d at 1512 (en banc) (“The test has nothing to do with what the best lawyers would have done. Nor is the test even what most good lawyers would have done. We ask only whether some reasonable lawyer at the trial could have acted, in the circumstances, as defense counsel acted at trial.”); see also Harich v. Dugger, 844 F.2d 1464, 1470 (11th Cir.1988) (en banc) (“It is not enough for petitioner to claim his counsel was ignorant of the Florida law. Petitioner must prove that the approach taken by defense counsel would not have been used by professionally competent counsel”); Provenzano v. Singletary, 148 F.3d 1327, 1332 (11th Cir.1998) (noting that counsel’s conduct is unreasonable only if petitioner shows “that no competent counsel would have made such a choice”); Burger, 107 S.Ct. at 3124 (in concluding that defense counsel’s not using character witnesses met reasonableness standard, Court pointed out that district court judge — presumably a reasonable lawyer— who heard the proffered mitigating evidence did not think it would have aided petitioner’s case).
VI. When courts are examining the performance of an experienced trial counsel, the presumption that his conduct was reasonable is even stronger.18 See Provenzano, 148 F.3d at 1332 (stating that “strong reluctance to second guess strategic decisions is even greater where those decisions were made by experienced criminal defense counsel” and that “[t]he more experienced an attorney is, the more likely it is that his decision to rely on his own experience and judgment in rejecting a defense” is reasonable); see also Burger, 107 S.Ct. at 3118 (reciting counsel’s impressive credentials in opinion finding that counsel rendered effective assistance).19
VII. In reviewing counsel’s performance, a court must avoid using “the distorting effects of hindsight” and must evaluate the reasonableness of counsel’s performance “from counsel’s perspective at the time.” Strickland, 104 S.Ct. at 2065. “[I]t is all too easy for a court, examining counsel’s defense after it has proved unsuccessful, to conclude that a particular act or omission of counsel was unreasonable.” 20 Id.; see Waters, 46 F.3d at 1514 (en banc) (“The widespread use of the tactic of attacking trial counsel by showing what ‘might have been’ proves *1317that nothing is clearer than hindsight— except perhaps the rule that we will not judge trial counsel’s performance through hindsight.”).
VIII. No absolute rules dictate what is reasonable performance for lawyers. Strickland, 104 S.Ct. at 2065 (“No particular set of detailed rules for counsel’s conduct can satisfactorily take account of the variety of circumstances faced by defense counsel or the range of legitimate decisions regarding how best to represent a criminal defendant.”); see also Roe v. Flores-Ortega, — U.S. -, 120 S.Ct. 1029, 1036-37, 145 L.Ed.2d 985 (2000) (rejecting bright-line rule that counsel “almost always” has duty to consult with defendant about appeal, reaffirming that the Court has “consistently declined to impose mechanical rules on counsel— even when those rules might lead to better representation”). “Any such set of rules would interfere with the constitutionally protected independence of counsel and restrict the wide latitude counsel must have in making tactical decisions.” Strickland, 104 S.Ct. at 2065; see also id. at 2066 (stating that “ rigid requirements [would] dampen the ardor and impair the independence of defense counsel, discourage the acceptance of assigned cases, and undermine the trust between attorney and client”); Waters, 46 F.3d at 1511 (en banc) (noting that Supreme Court has prohibited “[^Intensive scrutiny of counsel and [the creation of] rigid requirements for acceptable assistance”). The law must allow for bold and for innovative approaches by trial lawyers. And, the Sixth Amendment is not meant “to improve the quality of legal representation,” but “simply to ensure that criminal defendants receive a fair trial.” Strickland, 104 S.Ct. at 2065.
IX. Thus, no absolute duty exists to investigate particular facts or a certain line of defense. Under Strickland, counsel’s conducting or not conducting an investigation need only be reasonable to fall within the wide range of competent assistance. 104 S.Ct. at 2066 (stating that counsel “has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary”); cf. Kimmelman, 106 S.Ct. at 2588 (failure to file timely motion to suppress unlawfully obtained evidence amounts to constitutionally ineffective assistance when failure is based on counsel’s unreasonable mistake of law about the Government’s duty to supply certain information to defense counsel before trial) (emphasis added); Williams, 120 S.Ct. at 1514 (failure to conduct investigation ineffective because based, in part, on lawyer’s mistake of law that information not discoverable).21
*1318And counsel need not always investigate before pursuing or not pursuing a line of defense. Investigation (even a nonexhaustive, preliminary investigation) is not required for counsel reasonably to decline to investigate a line of defense thoroughly.22 See Strickland, 104 S.Ct. at 2066 (“[Strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation.”); id. (“In any ineffectiveness ease, a particular decision not to investigate must be directly assessed for reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s judgments.” (emphasis added)); Williams, 185 F.3d at 1236-37 (noting that this circuit has rejected idea that “strategic decisions can be considered reasonable only if they are preceded by a ‘thorough investigation’ ” and stating that, to be effective, counsel is not “required to ‘pursue every path until it bears fruit or until all hope withers’ ”); Rogers v. Zant, 13 F.3d 384, 387 (11th Cir.1994) (“By its nature, ‘strategy’ can include a decision not to investigate ... [and] a lawyer can make a reasonable decision that no matter what an investigation might produce, he wants to steer clear of a certain course.”); see also Holladay v. Haley, 209 F.3d 1243, 1252 (11th Cir.2000) (noting that our circuit has rejected “a per se rule of ineffective assistance where counsel does not consult family members”). For example, counsel’s reliance on particular lines of defense to the exclusion of others — whether or not he investigated those other defenses — is a matter of strategy and is not ineffective unless the petitioner can prove the chosen course, in itself, was unreasonable.23
X. Because the reasonableness of counsel’s acts (including what investigations are reasonable) depends “critically” upon “information supplied by the [petitioner]” or “the [petitioner’s own statements or actions,” evidence of a petitioner’s statements and acts in dealing with counsel is highly relevant to ineffective assistance claims. Strickland, 104 S.Ct. at 2066. “[An] inquiry into counsel’s eonver-*1319sations with the [petitioner] may be critical to a proper assessment of counsel’s investigation decisions, just as it may be critical to a proper assessment of counsel’s other litigation decisions.”24 Id. (“[W]hen a defendant has given counsel reason to believe that pursuing certain investigations would be fruitless or even harmful, counsel’s failure to pursue those investigations may not later be challenged as unreasonable.”).
XI. Counsel is not required to present every nonfrivolous defense; nor is counsel required to present all mitigation evidence, even if the additional mitigation evidence would not have been incompatible with counsel’s strategy. See Waters, 46 F.3d at 1511 (en banc) (noting that no absolute duty exists to present all possible mitigating evidence available: “Our decisions are inconsistent with any notion that counsel must present all available mitigating circumstance evidence.”). Considering the realities of the courtroom, more is not always better. Stacking defenses- can hurt a case. Good advocacy requires “winnowing out” some arguments, witnesses, evidence, and so on, to stress others. See Rogers, 13 F.3d at 388 (citing Jones v. Barnes, 463 U.S. 745, 103 S.Ct. 3308, 3313, 77 L.Ed.2d 987 (1983)); see also Waters, 46 F.3d at 1512 (en banc) (“There is much wisdom for trial lawyers in the adage about leaving well enough alone.”).
XII. No absolute duty exists to introduce mitigating or character evidence.25 See Tarver v. Hopper, 169 F.3d 710, 715 (11th Cir.1999) (noting that counsel is not “required to investigate and present all available mitigating evidence to be reasonable”) (citing Burger, 107 S.Ct. at 3126); Stanley v. Zant, 697 F.2d 955, 961 (11th Cir.1983) (no duty to present general character evidence); see also Waters, 46 F.3d at 1511 (en banc) (noting this court and Supreme Court have held counsel’s performance to be constitutionally sufficient when no mitigation evidence was produced even though it was available). See, e.g., Burger, 107 S.Ct. at 3126 (finding counsel effective even though counsel presented no mitigation evidence at all); Darden, 106 S.Ct. at 2474 (same).
These principles guide the courts on the question of “reasonableness,” the touchstone of a lawyer’s performance under the Constitution.
PERFORMANCE IN THIS CASE
Petitioner says that his trial counsel’s performance during the sentencing phase of his trial was unreasonable. Trial counsel at the sentencing phase called Petitioner’s mother and wife to testify, advanced two statutory mitigating factors, and stressed lingering doubt about Petitioner’s true guilt. Our court’s proper inquiry is limited to whether this course of action might have been a reasonable one. And, we begin with the strong presumption that it was. We conclude that — given *1320the record in this case and taking in the principles for ineffective assistance claims — Petitioner has failed, as a matter of law, to overcome the presumption.
Although Petitioner’s claim is that his trial counsel should have done something more, we first look at what the lawyer did in fact. Trial counsel focused on obtaining an acquittal and then, at sentencing, on lingering doubt.26 This defense was a reasonable one.27 We have said before that focusing on acquittal at trial and then on residual doubt at sentencing (instead of other forms of mitigation) can be reasonable. See Tarver v. Hopper, 169 F.3d 710, 715-16 (11th Cir.1999). Especially when — as in this case-— the evidence of guilt was hot overwhelming,28 we expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for relying on this seemingly reasonable strategy to defend his client.
*1321Trial counsel did not pursue character witnesses for mitigation;29 but he had other mitigators in hand. That trial counsel’s approach (preparing and presenting a case for doubt about Petitioner’s guilt instead of focusing on mitigating character evidence) was reasonable is even more clear in the light of the questionable value of the mitigating character evidence.
A lawyer reasonably could have déter-mined that character evidence would not be compelling in this case. And a lawyer reasonably could also fear that character evidence might, in fact, be counterproductive: it might provoke harmful cross-examination and rebuttal witnesses.30 Misgivings about hurtful cross-examination and rebuttal witnesses have been decisive to the Supreme Court when it determined that counsel was effective. See, e.g., Burger v. Kemp, 483 U.S. 776, 107 S.Ct. 3114, 3124-25, 97 L.Ed.2d 638 (1987) (concluding that failure to introduce character evidence was effective performance because witnesses could have been subjected to harmful cross-examination or invited other damaging evidence); Darden v. Wainwright, 477 U.S. 168, 106 S.Ct. 2464, 2474, 91 L.Ed.2d 144 (1986) (same); Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 2057 & 2071, 80 L.Ed.2d 674 (1984) (same). Trial counsel in this case has testified that he had these thoughts and concerns.
At the section 2255 hearing, trial counsel testified that, although he knew that testimony about Petitioner’s character would be admissible as mitigation evidence, he thought that character witnesses would not be very helpful or compelling, in this case. “[I]t would be at least questionable whether a sufficient impact of character type testimony could overcome a fixed opinion based on the other evidence ... [whether it] could change it from life to death. Or death to life.”31 The trial record indicates that counsel used other evidence and *1322stressed lingering doubt. He thought character evidence (even evidence of specific good acts by Petitioner) would not prevent the jury — if they were sure Petitioner had procured this murder — from giving Petitioner the death penalty.32 And trial counsel also questioned whether evidence of instances of Petitioner’s specific good acts would have been compelling, considering that the Government was not arguing that Petitioner was in all ways a bad man, but arguing that he had committed specific criminal acts, including offering to pay for a murder.33
As every reasonable trial lawyer knows, character witnesses that counsel called could be cross-examined by the Government. And as trial counsel said, such cross-examination might not be helpful to his case. It is uncontroverted that, based on his earlier interviews with people in the pertinent community, he knew that “some individuals in the community considered [Petitioner] to be a drug dealer” and “that there were people in the community [who] were afraid of him.”
Trial counsel also had seen at this very trial how a character witness’s testimony could be twisted by cross-examination and the arguments of opposing counsel. A witness at the guilt phase testified that Petitioner had given him some property for a house after the witness was newly married, even though the witness did not have the money to pay for the land. Trial counsel then had attempted to paint the story as good-act evidence. Trial counsel accurately noted that the Government, however, used this testimony to argue that Petitioner’s gift to this man was, in reality, part of a money laundering operation.34
That counsel’s concerns about using character evidence were reasonable is con*1323firmed by the transcript of the evidentiary hearing for section 2255 relief. At the hearing, the Government did effectively cross-examine the proffered character witnesses. The district court judge — the same, very experienced judge who presided at the murder trial itself — after seeing and hearing these witnesses, did not think they were helpful to Petitioner’s case because they were nullified on cross-examination. See Burger, 107 S.Ct. at 3124 (concluding that trial counsel acted reasonably in not calling witness at sentencing that district judge later heard fully at ha-beas hearing and found not to be helpful); see also White v. Singletary, 972 F.2d 1218, 1225 (11th Cir.1992) (questioning whether counsel would even have presented evidence had he possessed it because it had substantial internal weaknesses).
Trial counsel also testified that he was “fearful” of rebuttal witnesses: “I felt that the law enforcement community in Piedmont, in that part of the county, was hostile to [Petitioner], antagonistic to him. And that they certainly could have produced witnesses of that sort.”35 A reasonable lawyer could decide to limit reliance on character testimony instead of exposing the jury (right at sentencing) to a new string of Government witnesses who could testify to Petitioner’s bad acts.36 We must conclude that trial counsel’s approach to *1324the sentencing proceedings was a reasonable one.
Petitioner, on this record, has given us no cause to doubt this conclusion. Petitioner never testified at his section 2255 hearing. The reasonableness of a trial counsel’s acts, including lack of investigation or excluding character witnesses from the sentencing phase, depends “critically” upon what information the client communicated to counsel. Strickland, 104 S.Ct. at 2066. In this case, Petitioner and trial counsel shrouded the conversations between themselves in attorney-client privilege; 37 so we do not know to what extent Petitioner informed trial counsel’s acts.38 Therefore, given the absence of evidence in the record, we must assume counsel carried out his professional responsibility and discussed mitigation with his client.39 See Williams v. Head, 185 F.3d 1223, 1235 (11th Cir.1999). In addition, the section 2255 transcript is clear on two points: trial counsel testified — without dispute' — that he *1325frequently met with Petitioner before and during trial and that no one who spoke with trial counsel ever came forward with facts about character evidence that he thought would be helpful.40
In short, trial counsel, based on his professional judgment as an experienced trial lawyer, determined (or some reasonable lawyer could have) that his client had a fair chance for acquittal, saw (or some reasonable lawyer could have) character witnesses — with the potential dangers associated with cross-examination and rebuttal witnesses — as less than compelling in mitigation, and allocated (or some reasonable lawyer could have) his time and resources accordingly. Trials are full of imponderables. Nothing in the record indicates with force that this lawyer’s conduct was outside of the range of reasonable conduct.
When the guiding principles are applied to the record, the record will allow only one legal conclusion: Petitioner is due no relief.41 In the light of the *1326strong presumption that counsel was effective and the circumstances of the case, Petitioner has not met his heavy burden to prove that counsel’s acts — at sentencing, focusing largely on residual doubt and not investigating or presenting mitigating character witnesses' — were unreasonable.42 Nothing more needs to be said. The Constitution did not demand that trial counsel, *1327in this case, use more character witnesses.43
CONCLUSION
Petitioner’s evidence was insufficient to prove that his trial counsel’s acts were outside the wide range of professionally competent assistance. Thus, Petitioner has not met his burden under the law to prove ineffective assistance of counsel in this case. Because Petitioner cannot properly be granted relief, we affirm the district court’s denial of the writ.44
AFFIRMED.
COX, Circuit Judge, specially concurring, in which DUBINA, BLACK, HULL and MARCUS, Circuit Judges, join:
I join Judge Edmondson’s opinion in full. I write separately because I would also affirm the denial of Chandler’s § 2255 petition because Chandler has not shown prejudice from his counsel’s alleged deficient performance.
The district court assumed (without deciding) that Chandler’s counsel’s performance at sentencing had been deficient, but denied relief, concluding that Chandler had failed to establish prejudice. The district court’s ultimate conclusion as to prejudice answers a mixed question of law and fact, and we review that conclusion de novo, but we review the district court’s subsidiary factual findings only for clear error. See Strickland v. Washington, 466 U.S. 668, 698, 104 S.Ct. 2052, 2070, 80 L.Ed.2d 674 (1984).
The district court, after hearing the testimony of 27 character witnesses at the hearing on the ineffective assistance claim, found that the value of their testimony was severely undercut by three factors. First, most of the character witnesses’ knowledge of Chandler was “stale;” that is, their knowledge of Chandler’s good character pre-dated Chandler’s illegal activities, and the witnesses had little familiarity with him in the years leading up to the murder. Second, many witnesses also evidenced a lack of knowledge about Chandler’s character by testifying that they were unaware that, as the jury heard, Chandler bought, grew, and sold large quantities of marijuana. Third, the district court found that most of the character witnesses exhibited a strong bias for Chandler because they testified that, even if Chandler had committed certain bad acts,1 their opinions of Chandler would not have changed. As the district court found, “a witness’s high opinion of Chandler would have been of little moment to the jury if the witness believed that drug dealing and violent crimes were irrelevant to a person’s character.” United States v. Chandler, 950 F.Supp. 1545, *13281571 (N.D.Ala.1996). In summation, the district court found that “the mitigation evidence that Chandler’s trial counsel could have offered was of tenuous value.” Id.
*1327(1) Chandler's arrest in Georgia while attempting to purchase 100 pounds of marijuana;
(2) Chandler's flight from Georgia Bureau of Investigation Agent Skinner and Chandler's attempt to turn Skinner’s gun back on Skinner during a scuffle; and
(3) Chandler’s tape-recorded statement to a confidential informant that if he were “set up’’ again, he would have to kill someone.
*1328The district court’s findings about the value of this testimony are factual findings, subject to review only for clear error. A court’s factual finding is clearly erroneous only if “‘although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.’ ” University of Georgia Athletic Assoc. v. Laite, 756 F.2d 1535, 1543 (11th Cir.1985) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 542, 92 L.Ed. 746 (1948)). The district court made findings about the value of this testimony after seeing and hearing the witnesses, and those findings have support in the record. I cannot conclude that those findings are clearly erroneous.
Accordingly, in addressing the question of prejudice de novo, I give little weight to the character evidence that could have been introduced at sentencing. The ultimate question is whether Chandler has shown that any deficient performance prejudiced him such that, without the errors, there is a reasonable probability that the balance of aggravating and mitigating circumstances would have been different. See Bolender v. Singletary, 16 F.3d 1547, 1556-57 (11th Cir.1994) (citing Strickland, 466 U.S. at 687, 104 S.Ct. at 2064). Weighing anew the aggravating and mitigating factors, I note that the jury found two aggravating factors: (1) Chandler intentionally engaged in conduct intending that Shuler be killed and resulting in Shu-ler’s death, 21 U.S.C. § 848(n)(l)(C); and (2) Chandler procured Shuler’s killing by promising to pay something of pecuniary value, 21 U.S.C. § 848(n)(6).2 The two statutory mitigating factors, admitted by stipulation,were that (1) Chandler had no prior criminal record, 21 U.S.C. § 848(m)(6); and (2) the triggerman would not receive the death penalty, 21 U.S.C. § 848(m)(8). Chandler’s counsel also presented mitigating evidence in the testimony of Chandler’s wife and mother at sentencing.
On balance, whether Chandler was prejudiced is, as the district court noted, a close question. The jury had convicted Chandler of a particularly egregious crime. It does not appear to me, given the strong aggravating factors, that the addition of weak character evidence would have tipped the balance in favor of mitigation. I conclude, therefore, that the district court was correct in its determination that Chandler has failed to establish prejudice, and would affirm on that ground as well.
. We conclude that only Petitioner's claim of ineffectiveness at sentencing warrants more discussion. Therefore, for the reasons set out by the district court and in our earlier, but now-vacated, panel decision, we affirm the district court’s denial of habeas relief for Petitioner’s other claims.
. In March 1990, state law enforcement officers executed a search warrant at the home of one of Petitioner’s dealers, based on information provided by Shuler.
. Redden had tried over 1000 cases, had formerly been a prosecutor at the U.S. Attorney’s Office, was formerly president of the Alabama Bar, was a member of the American College of Trial Lawyers and the International Society of Barristers. He is listed in America's Best Lawyers for his criminal defense work. For further background, including academic honors, see the Martindale-Hubble Law Directory. By the way, the prosecutor said during the trial that Redden is an "extremely talented defense counsel, probably the best in the state [of Alabama].”
.At the section 2255 hearing, trial counsel testified that in his view "the guilt of murder to a capital degree ... was the weakest part of the [Government’s] case.” And he stated that, based on the circumstances of the Government’s case, he believed "there was a fair chance” his client would be found not guilty and it was even "less likely” that he would get the death penalty.
. In response to a question, trial counsel testified at the section 2255 hearing that, based on his "reasonable professional judgment,” he had not believed that a reasonable jury would impose the death penalty given the weak case against his client:
I did not feel that a reasonable minded jury would impose the death penalty on him, given not only the testimony as to what had occurred in November when Jarrell, Sr. undertook to kill Shuler, but also what occurred on the day that he did kill him. And that is, that they had consumed a tremendous amount of beer, been out there at Snow's Lake for a good while. I think they had given out of beer and gone back and gotten more beer. And that even if one assumed that Ronnie Chandler had offered him $500, well, this was still the act of a drunken mind overwhelmed by other things than a $500 offer.
And this view supports the argument (an argument mainly based on residual doubt about Petitioner’s responsibility for the shooting) trial counsel pressed at the sentencing phase.
. Trial counsel explained to the jury the mitigating nature of the women's testimony this way:
[The testimony of Petitioner’s wife and mother] was here to show that there was a life here that has had a stability to it, that has had some quality to it and I think that is apparent when you looked at those two ladies ... and the fact that here is a family that had tremendous stability.... He’s got three children, they are all by his wife. Here is a man who apparently has some skill of his hands who has worked in building his house and his parents' house, his brothers' houses and they've worked with him and this springs off the 80-acre farm that his father had with his father, way back. They built a sawmill, they cut trees, they made lumber, they collected rocks, they built houses and lives demonstrating lives with some purpose as opposed to life worthless. So that that is a mitigating factor that I think that you have certainly not just the right but the obligation to consider.
. Counsel also pointed out that Petitioner’s alleged statements to Jarrell about killing Shuler, if made at all, were made only twice — the first occasion Jarrell himself characterized as a joke. Counsel reiterated to the jury, "if it took place, it was a thing that was not precipitated.” The jury did not find extensive planning to he an aggravating factor.
. For example, witnesses testified that Petitioner had bought shoes for a boy who did not have any; had given money to a family to help them bury their son; had bought groceries for people he thought needed them; had bought dinner for members of his construction crew who did not have the money to buy their own; had given lunch money to a neighbor’s children; and had offered to allow a woman to stay at his house when her husband died.
.When cross-examined, the witnesses testified that knowledge of Petitioner’s drug dealing, attempt to turn a gun on a police officer, and statement about having to kill someone if he were set up again, would not change their opinion of Petitioner. The district court found that this testimony "largely nullified the persuasive value” of this character testimony because the jury would not be likely to credit witnesses who "believed that drug dealing and violent crimes were irrelevant to a person's character.”
.Most of these principles and presumptions were expressly set out in three ineffective assistance cases in which the Supreme Court specifically addressed the issue of ineffective assistance of counsel at the sentencing stage for failure to investigate and to present mitigating evidence. In each case, the Court determined counsel was not ineffective. See Burger v. Kemp, 483 U.S. 776, 107 S.Ct. 3114, 97 L.Ed.2d 638 (1987) (death penalty case); Darden v. Wainwright, 477 U.S. 168, 106 S.Ct. 2464, 91 L.Ed.2d 144 (1986) (death penalty case); Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) (death penalty case).
. There are different kinds of ineffective assistance claims. Here, Petitioner does not allege that his trial counsel’s performance was impaired by an incapacity (mental or physical), a conflict of interest, bad faith, or an unreasonable mistake of law. Therefore, we decide nothing today about these other kinds of cases.
. "The test for ineffectiveness is not whether counsel could have done more; perfection is not required. Nor is the test whether the best criminal defense attorneys might have done more. Instead the test is ... whether what they did was within the ‘wide range of reasonable professional assistance.' ” Waters, 46 F.3d at 1518 (en banc) (citations omitted).
. "It is important to note that judicial scrutiny of an attorney's performance is appropriately highly deferential because the craft of trying cases is far from an exact science; in fact, it is replete with uncertainties and obligatory judgment calls.” Bolender v. Singletary, 16 F.3d 1547, 1557 (11th Cir.1994).
. In accordance with this principle, courts must recognize that counsel does not enjoy the benefit of unlimited time and resources. See Rogers v. Zant, 13 F.3d 384, 387 (11th Cir.1994). Every counsel is faced with a zero-sum calculation on time,, resources, and defenses to pursue at trial.
And, a court must not second-guess counsel’s strategy. Waters, 46 F.3d at 1518-19 (en banc). By "strategy,” we mean no more than this concept: trial counsel’s course of conduct, that was neither directly prohibited by law nor directly required by law, for obtaining a favorable result for his client. For example, calling some witnesses and not others is "the epitome of a strategic decision.” Id. at 1512 (en banc); see also id. at 1518-19 (en banc); Felker v. Thomas, 52 F.3d 907, 912 (11th Cir.1995) (whether to pursue residual doubt or another defense is strategy left to counsel, which court must not second-guess); Stanley v. Zant, 697 F.2d 955, 964 (11th Cir.1983) (stating that reliance on line of defense to exclusion of others is matter of strategy).
. The presumption impacts on the burden of proof and continues throughout the case, not dropping out just because some conflicting evidence is introduced. "Counsel's competence ... is presumed, and the [petitioner] must rebut this presumption by proving that his attorney’s representation was unreasonable under prevailing professional norms and that the challenged action was not sound strategy.” Kimmelman v. Morrison, 477 U.S. 365, 106 S.Ct. 2574, 2588, 91 L.Ed.2d 305 (1986) (emphasis added) (citations omitted). An ambiguous or silent record is not sufficient to disprove the strong and continuing presumption. Therefore, "where the record is incomplete or unclear about [counselj's actions, we will presume that he did what he should have done, and that he exercised reasonable professional judgment.” Williams, 185 F.3d at 1228; see also Waters, 46 F.3d at 1516 (en banc) (noting that even though testimony at habeas evidentiary hearing was ambiguous, acts at trial indicate that counsel exercised sound professional judgment).
The presumption about which we write here is not some presumption that the particular defense lawyer in reality focused on and, then, deliberately decided to do or not to do a specific act. Instead, the presumption to which we refer is the presumption that what the particular defense lawyer did at trial — for example, what witnesses he presented or did not present' — were acts that some reasonable lawyer might do.
The Supreme Court has instructed that there is "a strong presumption that counsel's conduct falls within the wide range of reasonable professional assistance.” Strickland, 104 S.Ct. at 2065. This presumption is like the *1315“presumption of innocence” in a criminal trial, in which "the defendant is not required to come forward with proof of his innocence once evidence of guilt is introduced to avoid a directed verdict of guilty.” Black’s Law Dictionary 823 (6th. ed.1991). This presumption of competence must be disproved by a petitioner. Petitioner continually bears the burden of persuasion on the constitutional issue of competence and further, (adding the prejudice element) on the issue of ineffective assistance of counsel. See Strickland, 104 S.Ct. at 2064 (stating that “defendant must show that counsel’s performance was deficient” and that defendant must also show prejudice). Never does the government acquire the burden to show competence, even when some evidence to the contrary might be offered by the petitioner.
. To uphold a lawyer’s strategy, we need not attempt to divine the lawyer’s mental processes underlying the strategy. "There are countless ways to provide effective assistance in any given case.” Strickland, 104 S.Ct. at 2065. No lawyer can be expected to have considered all of the ways. If a defense lawyer pursued course A, it is immaterial that some other reasonable courses of defense (that the lawyer did not think of at all) existed and that the lawyer’s pursuit of course A was not a deliberate choice between course A, course B, and so on. The lawyer's strategy was course A. And, our inquiry is limited to whether this strategy, that is, course A, might have been a reasonable one. See generally Harich v. Dugger, 844 F.2d 1464, 1470-71 (11th Cir.1988) (en banc) (concluding — without evidentiary hearing on whether counsel's strategy arose from his ignorance of law— that trial counsel's performance was competent because hypothetical competent counsel reasonably could have taken action at trial identical to actual trial counsel), replacing vacated panel opinion, 813 F.2d 1082 (11th Cir.1987) (2-1 opinion) (remanding for evi-dentiary hearing on whether pursuit at trial of actual innocence defense, instead of intoxication defense or a combination of defenses, was informed strategic decision); Bonin v. Calderon, 59 F.3d 815, 838 (9th Cir.1995) (holding — where petitioner alleged that trial counsel’s mental processes were impaired by drug use — that, because an objective standard is used to evaluate counsel's competence, "once an attorney’s conduct is shown to be objectively reasonable, it becomes unnecessary to inquire into the source of the attorney’s alleged shortcomings”). See also Roe v. Flores-Ortega, - U.S. -, 120 S.Ct. 1029, 1037, 145 L.Ed.2d 985 (2000) ("The relevant question is not whether counsel’s choices were strategic, but whether they were reasonable.”).
We look at the acts or omissions of counsel that the petitioner alleges are unreasonable and ask whether some reasonable lawyer could have conducted the trial in that manner. Because the standard is an objective one, that trial counsel (at a post-conviction evidentiaiy hearing) admits that his performance was deficient matters little. See Tarver v. Hopper, 169 F.3d 710, 716 (11th Cir.1999) (noting that "admissions of deficient performance are not significant”); see also Atkins v. Singletary, 965 F.2d 952, 960 (11th Cir.1992) (“[Ineffectiveness is a question which we must decide, [so] admissions of deficient performance by attorneys are not decisive.”).
. If some reasonable lawyer might have not pursued a certain defense or not called a certain witness, we fail to understand why we would order a new trial on the ground that the actual lawyer had not used the defense or witness in the first trial: at the new trial, a different lawyer (even a reasonable one) might again not use the witness or defense. If two trials are identical, one should not be *1316constitutionally inadequate and the other constitutionally adequate.
. We accept that even the very best lawyer could have a bad day. No one's conduct is above the reasonableness inquiry. Just as we know that an inexperienced lawyer can be competent, United States v. Cronic, 466 U.S. 648, 104 S.Ct. 2039, 2050, 80 L.Ed.2d 657 (1984) (inexperienced does not mean ineffective), so we recognize that an experienced lawyer may, on occasion, act incompetently. Our point is a small one: Experience is due some respect.
. ''[Counsel] had been practicing law in Wayne County for about 14 years, had served as the county’s attorney for most of that time, and had served on the Board of Governors of the State Bar Association. About 15 percent of his practice was in criminal law, and he had tried about a dozen capital cases. It is apparent that he was a well-respected lawyer, thoroughly familiar with practice and sentencing juries in the local community.” Burger, 107 S.Ct. at 3118.
.For example, "[i]t is common practice for petitioners attacking their death sentences to submit affidavits from witnesses who say they could have supplied additional mitigating [] evidence, had they been called or ... had they been asked the right questions.” Waters, 46 F.3d at 1514 (en banc). But "[t]he mere fact that other witnesses might have been available or that other testimony might have been elicited from those who testified is not a sufficient ground to prove ineffectiveness of counsel.” Id. (noting that such witnesses show nothing more than that, "with the luxury of time and the opportunity to focus resources on specific parts of a made record, post-conviction counsel will inevitably identify shortcomings”). And, basing the inquiry on whether an investigation (if one had been undertaken) would have uncovered mitigating evidence (or witnesses) is an example of judging counsel’s acts from the benefit of hindsight. The proper inquiry was articulated in Rogers v. Zant: "Once we conclude that declining to investigate further was a reasonable act, we do not look to see what a further investigation would have produced.” 13 F.3d 384, 388 (11th Cir.1994).
. We do not read Williams to declare a per se rule of law that a defense lawyer must present character witnesses at the sentencing phase or that a defense lawyer (no matter what his client may have informed or instructed him) must in every case investigate purely to see if character witnesses might exist who might be of help at the sentencing phase.
We understand Williams to create no mechanistic rule of law at all for investigation or for presentation of evidence in capital cases. See generally Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 399, 5 L.Ed. 257 (1821) (Marshall, C.J.) ("It is a maxim not to be disregarded, that general expressions, in every opinion, are to be taken in connection with the case in which those expressions are used.”); Crawford-El v. Britton, 523 U.S. 574, 118 S.Ct. 1584, 1590, 140 L.Ed.2d 759 (1998) (“There is, of course, an important difference between the holding in a case and the reasoning that supports that holding.”). In particular, we do not understand the Supreme Court to make the ABA Standards part of the highest law of the land, even if one accepts that those standards reflect a good policy. We remember that these ABA Standards were also mentioned in Strickland; but the Court went on to shun per se rules. Strickland, 104 S.Ct. at 2065; accord Roe, 120 S.Ct. at 1036-37.
We read Williams to decide this question about lawyer performance: where a capital defendant (with a significant criminal record) has repeatedly admitted before trial that he, in fact, did kill the victim in the course of a theft so that a defense of factual innocence hardly existed and where, at the sentencing phase, the defense counsel, while repeatedly telling the jury that it was hard to explain why the jury should spare defendant's life, presents no evidence of the defendant's "nightmarish” childhood (criminal neglect by his parents, repeated severe beatings by his father, and abuse in a foster home) — not be*1318cause defense counsel thought such evidence would not be compelling (defense counsel testified at a habeas hearing that it was important evidence which he would have used had he known of it), but because he "incorrectly thought” that state law barred his access to the juvenile and social services records in which such information could be found — the defense lawyer’s performance at sentencing was deficient.
But the present case is different from Williams not only in the kinds of evidence that are involved, but in its other material facts: for example, here defendant's factual guilt was strongly disputed (and still is); here no one contends that the defense counsel’s not presenting more character witnesses was due to a mistaken view of the law (he did not think a search for such witnesses was prohibited by law or that character witnesses would, as a matter of law, be excluded); here defense counsel has never said that character witnesses would be very helpful or that he would have used them if he had known of them (and the trial judge who has heard and seen the pertinent character witnesses has said that, in his opinion, they would not have been helpful if introduced); here defense counsel did raise mitigating factors unavailable in Williams.
Williams cannot command the outcome for this case; the cases’ facts are materially different, allowing different outcomes under Strickland.
. As we have recognized, Strickland’s approach toward investigation "reflects the reality that lawyers do not enjoy the benefit of endless time, energy or financial resources.” Rogers, 13 F.3d at 387. How a lawyer spends his inherently limited time and resources is also entitled to great deference by the court. See White, 972 F.2d at 1224 ("[G]iven the finite resources of time and money that face a defense attorney, it simply is not realistic to expect counsel to investigate substantially all plausible lines of defense. A reasonably competent attorney often must rely on his own experience and judgment, without the benefit of a substantial investigation, when deciding whether or not to forego a particular line of defense....”).
. Requiring that counsel always do certain acts to be found effective (for example, interviewing some of petitioner’s neighbors for mitigation evidence) would contravene the Supreme Court’s directive that no set of detailed rules for counsel’s conduct should be used to evaluate ineffectiveness claims. See Strickland, 104 S.Ct. at 2065.
. And, when the circumstances of a claim make these conversations relevant, the petitioner can rarely (if ever) satisfy his burden to disprove the presumption of effective assistance without disclosing the substance of these attorney-client conversations. Cf. Laughner v. United States, 373 F.2d 326, 327 (5th Cir.1967) (refusing to allow petitioner, who requested a section 2255 evidentiary hearing, to invoke attorney-client privilege to "eliminate the one source of evidence likely to contradict his allegations”); Williams, 185 F.3d at 1235 ("Given the lack of clarity of the record, we presume that [counsel] talked with [petitioner] as part of his effort to ascertain whether there was any mitigating circumstance evidence....”).
. While Petitioner is correct that capital defendants have a right to present just about any evidence in mitigation at the sentencing phase, this right is the right to be free of governmental interference with the presentation of evidence. See generally Hitchcock v. Dugger, 481 U.S. 393, 107 S.Ct. 1821, 1824, 95 L.Ed.2d 347 (1987). The question of admissibility of evidence is different from whether counsel acted reasonably in not introducing the evidence. See Burger, 107 S.Ct. at 3123 n. 7. And the cases concerning the constitutional right of defendants not to be precluded or limited by the state or the court in their presentation of mitigation evidence at sentencing do not support the proposition that, if counsel does not present all possible mitigation at sentencing, then defendant has been denied some constitutional right.
. In his sentencing argument, defense counsel did not use the words 'lingering doubt” or "residual doubt.” But, as the government pointed out in its first brief filed in this court, defense counsel's argument at sentencing attacked the government’s aggravating factors: the factors expressly included that Petitioner "intentionally engaged in conduct intending that [the victim] be killed and resulting in [the victim’s] death.” And, trial counsel did argue at some length that the evidence showing that defendant in fact had caused the actual killer to shoot the victim was disputed evidence; and he pointed to the weakness of the evidence as a ground for a sentence other than death: for example, the lawyer argued "What prompted [the killer] in his actions. How much did anything that was said to him on that day by [Petitioner] impel him, motivate him to do what he did after twenty-three beers on that date, twenty-three beers before he shot the man."
In the context of the trial and sentencing proceeding, defense counsel’s argument, stressing the lack of strong evidence of guilt, cannot be said to be unconnected to "lingering doubt.” We recognize the argument as a lingering-doubt argument and would not approve a district court’s finding otherwise.
. At least when guilt in fact is denied, a "lawyer’s time and effort in preparing to defend his client in the guilt phase of a capital case continues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715 (11th Cir.1999); see also Darden v. Wainwright, 477 U.S. 168, 106 S.Ct. 2464, 2473, 91 L.Ed.2d 144 (1986) (rejecting petitioner's argument that counsel had only spent the time between conviction and sentencing preparing the case for mitigation because "counsel engaged in extensive preparation prior to trial, in a manner that included preparation for sentencing ”) (emphasis added); Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 1769, 90 L.Ed.2d 137 (1986) ("[I]t seems obvious to us that in most, if not all, capital cases much of the evidence adduced at the guilt phase of the trial will also have a bearing on the penalty phase. ...”).
Trial counsel, at the section 2255 hearing, testified that he had believed "there was a fair chance” his client would be found not guilty, and it was even "less likely” that he would get the death penalty given the circumstances of the murder case. These views are reasonable considering the evidence at the trial.
In this case, when we refer to trial counsel's testimony explaining his personal mental processes (assessing the strengths of the prosecution’s case, opining on the value of character witnesses and so on), we are not accepting that his words represent his heartfelt views, that is we are not crediting his testimony as absolutely true; but we point to this lawyer’s testimony as illustrating the kinds of thoughts some lawyer in the circumstances could — we conclude — reasonably have had. The trial counsel’s testimony is not essential to today’s affirmance.
.We have accepted that residual doubt is perhaps the most effective strategy to employ at sentencing. See Tarver, 169 F.3d at 715-16 (citing law review study concluding that "the best thing a capital defendant can do to improve his chances of receiving a life sentence ... is to raise doubt about his guilt”). Counsel cannot be held to be ineffective when he has taken a line of defense which is objectively reasonable.
The jury in this case was instructed at the guilt phase this way: "[I]t is not necessary that the defendant's guilt be proven beyond all possible doubt. It is only required that the Government’s proof exclude any reasonable doubt concerning the defendant’s guilt.” As this instruction shows, the law recognizes that jurors who have found a person guilty of a crime may well still have doubt about his true guilt. Thus, the law itself points to and lays the foundation for a good argument based on lingering doubt when the jury is later asked to impose death, the ultimate and most irremediable punishment. Nothing about this argument signals submissiveness or fatalism; stressing residual doubt is a straightforward and sound defense.
. Petitioner tries to make much of trial counsel’s asking Petitioner's wife — after the guilty verdict — to find witnesses and of trial counsel’s saying at the section 2255 hearing that he had a hope that some character witnesses might be presented to him (though he stated that it was not a hope “that I really spent a lot of time or effort on or felt that there was that much time to spend on’’). Petitioner claims this request of his wife shows that trial counsel really wanted character witnesses but had not invested the time and energy to find them. This circumstance, even if true, would not be decisive: recall especially that our inquiry is an objective one and that we are not attempting to grade this counsel’s performance, but are looking at the trial just to see if the trial (as it was) was adequate. Moreover, the evidence of a request to the wife — at least, as strongly — shows the laudable fact that trial counsel, like most good trial lawyers, was flexible and opportunistic: he did not think character witnesses would be helpful or a good use of his time to pursue, but a reasonable lawyer might not foreclose himself altogether from considering some if they were presented to him. In fact, one such character witness was presented to trial counsel; counsel did not use him.
. Two character witnesses were presented at the sentencing stage: Petitioner’s wife and mother. Petitioner also argues that trial counsel was ineffective for his limited preparation of and for his limited questioning of the wife and mother. Petitioner contends that, if trial counsel had asked the right questions, the lawyer could have elicited more mitigation testimony from them. But see Waters v. Thomas, 46 F.3d 1506, 1514 (11th Cir.1995) (en banc) (counsel not ineffective for failing to elicit more testimony from the witnesses because perfection is not required). Considering concerns about opening the door, a reasonable lawyer might well limit his questioning of the wife and mother: as it was, the wife and mother did escape cross-examination altogether.
The testimony that was elicited reminded the jury that Petitioner was a husband, a father, and a son: humanizing circumstances. The wife and mother addressing the jury also reminded the jury that innocent people would suffer if this man was put to death. Bringing the family’s existence and pain to the attention of the jury is powerful in and of itself. See id. at 1519 (noting that, sometimes, just having a witness on the stand can humanize petitioner in eyes of jury). We cannot say the Constitution demanded that more be done with these witnesses.
.Trial counsel testified:
Q: And that specific instances of the defendant helping people is probably the most compelling of that type of evidence, a mere *1322opinion that he’s a good person is one thing, but to say he did this for me is even better. You recognized that, did you not?
A: Well, I recognize its admissibility as a compelling nature, I wouldn’t test it.
Q: You didn’t think that type of testimony would be very helpful, is that your testimony?
A: Well I'm — if that’s your question, I say in the whole picture of things as it existed at that time, if that jury was going to give him the death penalty based on what they had heard, their minds were made up as to that, I doubted that compelling would be an appropriate word.
. Trial counsel stated that he did not need to know what a witness would say to determine whether the witness would be compelling at mitigation. He stated that you assume what a piece of testimony might be and “assume the most favorable testimony that you might get and then form some judgment, not the most reliable judgment in the world, but some judgment about how compelling it might be." We agree.
. The record shows:
[Habeas Counsel] Q: ... [Instances like that [of good acts], would have been something to refute the government's claim that [Petitioner] was just a bad person, wouldn’t it?
[Trial Counsel] A: If that had been the claim, yes.
Q: All right. But they did claim that, did they not?
A: Well, I don’t think they claimed that he was just a bad person. I think that they claimed that he was guilty of a specific offense with which I disagreed, but I can't say that the case was just that he was a bad person.
.To anyone familiar with popular books and films about organized crime, such as Mario Puzo's The Godfather, the idea that a criminal was helpful to many people in his community (perhaps on account of his goodness or perhaps out of maintaining the very helpful goodwill of the community in which the defendant is operating a criminal enterprise) would be familiar. In fact, the generosity of a man, convicted of being a drug kingpin, might reenforce those stereotypes. For a lawyer to worry about counterargument even to evidence of good deeds is not unreasonable.
The record from the evidentiary hearing provides other examples of how the character evidence could be considered aggravating. One man testified that Petitioner had donated $10,000 to the church, but the Government brought out on cross that Petitioner's entire reported income for the year was $10,000. And even those who testified that Petitioner was a regular churchgoer, admitted on cross that his attendance had dropped off in the years preceding his arrest. See Stanley v. Zant, 697 F.2d 955, 970 (11th Cir.1983) (questionable whether evidence would have been perceived as mitigating by the jury because church attendance could be perceived as aggravating: "if [he] went to the church, then he should have known the extreme culpability of his conduct”).
. Trial counsel testified that deciding whether to put on character evidence requires a balancing determination, which he made in this case, and that he "[had] to evaluate the impact of that along with everything else that you evaluate.” The record contains the following exchange:
Q: And I believe you stated that in determining whether or not to put on a character witness in view of the possibility of cross-examination by the government concerning various bad acts requiring a balancing act on your part, to determine whether or not the value of the witness, character witness would outweigh that possible damage done by that cross examination; is that correct?
[Trial Counsel] A: Well, balancing of the factors favorable and unfavorable and what I am saying is that that normally doesn't back you off putting on a character witness, but it’s a fact that they might be asked have you heard about so and so.
Q: Did you perform this balancing in this case?
A: I would think more of what might have been produced and I think I had said that not just on cross-examination of witnesses. [] But by rebuttal witnesses.
. Petitioner argues that counsel's worries about opening the door on cross-examination and the presentation of rebuttal witnesses were unreasonable because negative information about Petitioner had come out earlier at the guilt stage of the trial. Apart from the point that the timing of when evidence comes in may be as important as the evidence itself, we make these points.
We note the Government, on appeal in response to questioning, pointed out two examples of evidence from the section 2255 hearing, not introduced at trial, which might have come out at sentencing: (1) Petitioner had given a .357 magnum handgun to his fifteen-year-old son, and (2) the testimony of a witness (Scottie Surrett) — kept out at the guilt phase by Petitioner's trial counsel — implicating Petitioner in the murder of a missing “dope stealer.”
More important, Petitioner, who bears the burden in this case, never presented evidence that the fears of trial counsel about hurtful rebuttal witnesses were imaginary and baseless. For example, Petitioner’s habeas counsel, at the evidentiary hearing, could have asked trial counsel precisely what trial counsel feared and why, but the questions were not asked. We must not just assume that defense counsel's worries were baseless. For all we know, trial counsel actually discussed this issue with his client; and the client informed him of potentially harmful witnesses or information.
Furthermore, even a competent trial lawyer may be unable to articulate exactly what cross-examination and rebuttal witnesses he fears because the scope of discovery in federal criminal cases is limited. See, e.g., United States v. Fischel, 686 F.2d 1082, 1090 (5th Cir.1982) ("Discovery in criminal cases is narrowly limited.”). But the lack of articulation does not make his performance at trial incompetent. The fear of the unknown may, itself, be reasonable. For a defense lawyer to be a bit risk averse in a capital case is no indicator of incompetence. And recall in this case, trial counsel had (and did use) other mitigators; so, it was not a case of his having only one way to go. Compare Williams, 120 S.Ct. at 1514 (deeming counsel incompetent because failure to introduce voluminous mitigation evidence, even if it contained some unfavorable evidence about defendant’s juvenile record, was unreasonable when only al*1324ternative was to rely on the mitigating value of a voluntary confession to brutal murder and when counsel's mistake of law impacted on the omission of other evidence).
.Trial counsel censored himself; and given Petitioner's objections, the district court shielded the attorney-client conversations. At the 2255 hearing, the Government asked trial counsel if he had discussed Petitioner's arrest in Georgia with Petitioner. The following colloquy then occurred:
[Trial Counsel]: I’m aware of Your Hon- or’s order with reference to waiver of a confidentiality and this is not a—
[Habeas Counsel]: Your Honor, I would, for the record object to going into any attorney/client conversations in light of the fact that we have not gone into any of those on the direct so I think it’s beyond the scope of direct and I do not think that any of our examination touched on those issues which would constitute a waiver of the attorney/client privilege. Our examination dealt solely with his investigation with regard to other witnesses and with regard to specific particular witnesses. Not anything with regard to conversations with his client. I haven’t gone into that one whit. And I don't think I’ve waived the privilege in that regard nor do I believe that it is within the scope of the direct examination.
[U.S. Attorney]: Your Honor, the very basis of their allegations deal with what [trial counsel] knew or should have known at the time that this trial took place. And I think very clearly what his client told him or may have advised him concerning this information is relevant to cross-examination.
The Court: I think this is going to be an area that on a situation by situation, case-by-case basis, we may have to make rulings. (Emphasis added). The district court then sustained habeas counsel’s objection. And trial counsel, from the stand, at the end of this exchange told the judge: "I’m not raising an objection. I'm going to do what the court directs. But I want to be sure where we are. I think that there might be situations in which an answer to a question like that might implicitly indicate what was said.”
. The judge disallowed the substance of conversations based on attorney-client privilege and said he would make further rulings on an instance-by-instance basis. Habeas counsel argues that he had only objected to issues outside of the scope of the hearing. We observe, however, that an objection was made and sustained on whether counsel discussed a piece of evidence with his client even though Petitioner was challenging trial counsel’s failure to investigate sufficiently the author of this piece of evidence for trial. We also observe that trial counsel was careful himself not to touch on the conversations with his client. We find it troubling when a petitioner thinks he can meet his burden to show ineffectiveness even though he shielded and avoided pertinent conversations which would allow the courts to assess what trial counsel had learned from his client.
Habeas counsel claims he did not shield Petitioner’s conversations with the trial counsel about mitigation; habeas counsel says that no one asked about it. Even if habeas counsel had not raised an objection to shield the pertinent conversations, we stress he should have asked about them because Petitioner bears the burden of showing that trial counsel’s acts were unreasonable. And, as Strickland says, what the client said and did is critical to that proof.
. The effect of attorney-client privilege was argued by the Government in its briefs, and both sides addressed this issue at oral argument. This issue was sufficiently argued, see FSLIC v. Hardison, 813 F.2d 370, 373 n. 3 (11th Cir.1987) ("briefs are read liberally to ascertain the issues raised on appeal”), especially considering that we make no new law on this issue, but merely apply the existing case law from this circuit.
. Trial counsel (without contradiction) testified: "what I was preparing to do was to try to defend the case and there was not volunteered at that — in any of that among the people I talked with or presented to me anything that I considered would be helpful on a sentencing phase.” And "nobody had ever come forward and said such things [examples of good acts] and the only example of that type of thing that we had was in evidence.” He stated further, "nothing was volunteered to me that I considered of value in there.” See generally Collins v. Francis, 728 F.2d 1322, 1349 (11th Cir.1984) (counsel not ineffective for not investigating witnesses in mitigation when defendant failed to alert counsel to their existence). See also Strickland, 104 S.Ct. at 2066 ("[W]hen a defendant has given counsel reason to believe that pursuing certain investigations would be fruitless or even harmful, counsel’s failure to pursue those investigations may not later be challenged as unreasonable. ”).
. It certainly would have been less work for us to remand. And, we understand that appellate courts should not resolve genuine issues of disputed facts when those facts are material, in the light of the applicable law, to the outcome of the case; so, we do not do that. But, in our view, if we take seriously the principles that we have set out in this opinion, especially those dealing with objective reasonableness, no need for a remand arises. About remand, this principle of law is particularly important: We are not reviewing the quality of the specific lawyer's judgment processes that underlie his conduct at trial. (We must remember that Strickland passed through this court on its way to the Supreme Court and that some of our judges — who believed that the lawyer's performance in Strickland was inadequate under the Constitution— said they held that view because they were not convinced that the particular trial lawyer in that case had actually made a thought-out decision; the Supreme Court, however, determined the lawyer to be effective applying an objective standard. Washington v. Strickland, 693 F.2d 1243, 1283-84 (5th Cir. Unit B 1982) (en banc) (Johnson, J., concurring in part and dissenting in part), rev’d by 466 U.S. 668, 104 S.Ct. 2052, 2070-71, 80 L.Ed.2d 674 (1984).) We are reviewing the lawyer's conduct at trial and asking only whether some reasonable lawyer could have acted that way. In the light of the legal standards for lawyers' performance, we know enough to decide this case. We cannot honestly remand when, if the district court concluded that the trial lawyer’s performance was deficient, we already know that we would reverse that judgment because it would be contrary to law, given the record.
The record in this case — even when read in Petitioner’s favor — presents this legal question: Has a defendant proved his lawyer’s performance to be deficient — that is, totally beyond the outside border of the "wide range” of reasonable performance — at sentencing when the defendant (with no prior criminal record) is convicted of procuring a murder and his lawyer has invested most of his time and energy in defending against conviction; when the evidence of guilt is not overwhelming, relying largely on the testimony of the actual killer who has told a variety of stories and who has been promised that he will not be executed; when defense counsel did present witnesses as well as other evidence in mitigation; when defense counsel did not present other character witnesses who would have testified to his past good works; when no one knows what instructions and information about the availability and use of character witnesses were supplied by defendant to defense counsel; when defense counsel has never said that the pertinent character witnesses would be compelling or that he would have used them; when nothing indicates that defense counsel's act in not presenting more character witnesses was caused by his having some erroneous view of the law; when defense counsel argued to the juiy. at sentencing that the evidence that defendant *1326actually prompted the admitted killer to do the killing was in dispute (an argument that we say must be seen as a lingering-doubt argument given the context); when defense counsel stressed the lack of a criminal record for his client and that the actual killer would not be executed?
As we understand the law, we must conclude the answer is "no.” Considering these circumstances, we believe that some lawyer could reasonably have not presented the character witnesses, although we also accept that other lawyers could reasonably have presented the witnesses.
To us, the basic disagreement in this case among the court's judges is not about facts— what happened, happened — but about law and what kinds of facts are legally important given the objective legal standard. In a less academic sense, our disagreement also reminds us that lawyers (for we are lawyers, too) can easily disagree on almost any point about how to proceed in a trial and that, because trying cases is no exact science, no level of skill or excellence can exist at which a trial lawyer (who has been unsuccessful in a case) can be removed from intelligent criticism in hindsight.. There is no end to what might have been done differently.
. The district court, having concluded that no prejudice was proved in this case, did not specifically address whether counsel’s performance was deficient. But no remand is required in this case given (1) the developed record complete with an evidentiary hearing (which Petitioner does not argue was inadequate and at which no one testified about what Petitioner had said to his trial counsel about the benefit and danger of using character witnesses), (2) the strong presumption that counsel's acts at trial were acts that a reasonable lawyer could have taken, and (3) the reality that Petitioner bears the heavy burden of proof on ineffectiveness. Remand is not required when the record is sufficiently developed and a "complete understanding of the issues is possible in the absence of separate findings.” Tejada v. Dugger, 941 F.2d 1551, 1555 (11th Cir.1991) (considering habeas issue not addressed by district court); see Black, et ah, Federal Appellate Procedure— 11th Circuit, § 12:189 (1996) ("The trial court's failure to make necessary findings of fact constitutes harmless error when the facts necessary to the judgment have been incontestably demonstrated based on uncontrovert-ed evidence.”); see also Strickland, 104 S.Ct. at 2070-71 (Court applying different legal standard than applied by lower courts to record and declaring — without remand for more factfinding — defense counsel’s conduct not to be ineffective), rev’g, 693 F.2d 1243 (5th Cir.1982) (en banc) (remanding to district court because more district court findings needed to reach conclusion on ineffective assistance issue); Harich v. Dugger, 844 F.2d 1464, 1470 (11th Cir.1988) (en banc) (deciding question of ineffectiveness without remanding for district court factfindings). And we note that, neither party, in the briefs, contended that the record is insufficient and requires remand.
Moreover, courts have decided ineffective assistance claims on direct appeal, without the benefit of district court factfinding. See, e.g., United States v. Shukri, 207 F.3d 412, 418 (7th Cir.2000) (denying ineffective assistance of counsel relief on direct appeal even though issue was not raised to district court below because both parties requested it and claim of incompetence could be rejected considering record of trial); United States v. Fortson, 194 F.3d 730, 736 (6th Cir.1999) (on direct appeal, deciding no deficient performance because record was adequately developed to allow the court to assess the merits of the issue, government did not contend that record was insufficient, court concluded further factual development was unnecessary and because court could "conceive of numerous reasonable strategic motives” for counsel’s actions at trial (emphasis added)). And any alleged lack of evidence in this record necessitates neither a remand nor a ruling for Petitioner, but must be considered in the light of Petitioner’s burden to overcome the presumption of competence, that is, to show ineffective assistance. Cf. United States v. Montes-Mercado, 2000 WL 623143 (9th Cir.2000) (unpublished opinion) (refusing to conclude counsel ineffective for failing to provide reasons for his acts at the evidentiary hearing in part because petitioner never asked for an explanation); United States v. Torres, 845 F.2d 1165, 1172 (2d Cir.1988) (rejecting ineffective assistance claim, not considered by district court, as matter of law because proffered evidence of deficient performance did not "rise to the level necessary to overcome the strong presumption that counsel’s performance was reasonable under [Strickland]”).
. Never do we even hint that this course was either the best way or the only reasonable way to proceed in this case. To decide the constitutional question, we do not need to decide such things.
. About the question of prejudice, the court decides nothing. We need not. But several judges would be inclined to affirm on account of no prejudice, even if trial counsel’s performance was deficient.
About today’s dissenting opinions, we refrain from commenting, except that some of them bring to mind the words of Justice Jackson:
There has been much undiscriminating eulogy of dissenting opinions. It is said they clarify the issues. Often they do the exact opposite. The technique of the dissenter often is to exaggerate the holding of the Court beyond the meaning of the majority and then to blast away at the excess. So the poor lawyer with a similar case does not know whether the majority opinion meant what it seemed to say or what the minority said it meant.
Robert H. Jackson, The Supreme Court in the American System of Government, 18-19 (1995).
. These included:
. The Government presented evidence that Chandler committed the murder after substantial planning and premeditation, 21 U.S.C. § 848(n)(8), but the jury rejected that factor.